This invention relates to a computer system in which a cluster is constituted by multiple nodes in each of which a virtual server is operated, and more particularly, to a technique of upgrading system software.
JP 2005-148922 A discloses a technique of upgrading system software of a node. According to this technique, it is possible to suppress the downtime of a service provided to a client computer (client PC) as much as possible. Specifically, after the service provided to the client PC is failed-over to a standby node, the system software of the node is upgraded. With this operation, the client PC can use the service even while the system software of the node is being upgraded.
JP 2004-227127 A discloses a computer system in which multiple virtual servers are operated in a node. In this computer system, each virtual server provides a service. Computer systems of this type include a computer system in which a virtual server is operated for each application program group. In this computer system, a root filesystem of the virtual server is changed according to a change of environment. With this operation, different settings can be implemented in individual virtual servers. The root filesystem includes system software and setting information which is necessary to operate the system software.
JP 2005-267327 A discloses a technique of improving the reliability of a computer system in which multiple virtual servers are operated in a node. In this computer system, a filesystem of each virtual server is shared by multiple nodes which constitute a cluster. The virtual server is operated in one of the multiple nodes which constitute a cluster. During failover, a virtual server operated in the failover source is shut down. The virtual server which has been shut down is booted in a failover-destination node. With this operation, the multiple nodes which constitute a cluster can continuously provide services to the client PC.